An Effective Intrusion Detection System using Enhanced Multi relational Fuzzy Tree

Abstract

Today, the enormous development of computer networks and communication technology, we rely heavily on network connections. The substantial consumption of social networks and internet leads to the drastic increment in the data which are naturally complex and sparse too. The data are stored in multiple database relations associated with primary and foreign keys.  The internet attack is a main type of issue in computer networks. Numerous Network Intrusion Detection Systems (NIDSs) have been designed based on traditional data mining methods to identify and ease the network attacks. But the methods were suitable for single relational data.  This paper proposes a novel method for classifying KDD CUP 99 intrusion detection data using Enhanced Multi relational Fuzzy decision tree (EMRFT). The generated tree is optimized based on genetic approach. The outcomes of empirical analysis show that the EMRFT achieves high prediction performance and less induction time in classifying network intrusion.